EDPFRS: ENHANCED DYNAMIC POPULAR FILE REPLICATION AND SCHEDULING FOR DATA GRID ENVIRONMENT

A data grid is a structural design or cluster of services that enables transmission of a huge amount of geographically distributed data. Hence, it requires massive storage resources to store enormous data files. Data replication provides a solution for efficiently managing the data files in the huge distributed grid environment. It aids in enhancing the data availability and reducing the overall access time of the file. This paper presents an algorithm called Enhanced Dynamic Popular File Replication and Scheduling (EDPFRS) to enhance the data replication and scheduling of a file based on its popularity and the temporal localities in a data grid environment efficiently. For which a connectivity graph is created for a set of dependent files in each grid site to find the frequently accessed file replicas and maintain the consistency of the updated replicas. The appropriate replica file is located using the parameters such as file access rate and request for accessing the file. A Two-fold scheduling policy is applied to both the master node and head node to reduce the overall data access time in the grid. The best file replica is chosen based on the parameters such as network bandwidth, load rate, and computational capacity of the node. The proposed EDPFRS algorithm excels in replicating and scheduling the files in a grid environment